What is Bluescape AI?
Bluescape AI is a platform that can be used to seamlessly consolidate complex, multi-step workflows and diverse applications into a unified, secure, and collaborative hub for generative AI innovation.
Whether you are a developer or a large enterprise, Bluescape’s API is a robust, extensible architecture to create bespoke workflows in a virtual workspace. The API supports round-trip exchange of data, files, and metadata between Bluescape workspaces and external sources, including digital asset management systems and cloud storage providers. Additional capabilities include model fine-tuning and the ability to bring your own AI.
As part of of our Early Access program, you can enjoy access to a pre-built environment where you can experience just some of what is possible in Bluescape!
How to use your early access workspace
Some Bluescape Basics
If this is your first time using Bluescape and you need some tutorials on how to navigate and interact with your workspace, visit our Getting Started pages in our community for various guides on how to navigate, add content, use our robust collaborating capabilities, interact with images, and more!
Image Prompting
When you first join your workspace, there will already be a prompting form ready for you to start generating images. There are many modifications you can make to outputted images:
Include: Where you insert the instructions for the generative AI engine to create your images.
Do not include: Also known as “negative prompts”, you can insert specific things that you want to exclude from the images.
Variations: How many images you want it to generate (9 is currently the maximum).
Size: The overall dimensions of the outputted images, namely Square, Wide/Landscape, and Tall/Portrait options.
Style: The particular aesthetic and visual characteristics of the image, ranging from Photographs to 3D Renders.
Camera: What kind of camera the image appears to be shot from. Our favorite is Fujifilm 
Angle: The position or perspective from which a photograph is taken. It determines the viewing angle and composition of the image.
Lens: Plays a crucial role in capturing and shaping the incoming light, affecting factors such as the angle of view, depth of field, and overall image quality.
Speed versus detail: The balance between the generation speed and the level of detail in the outputted images. The further the slider is to the left, the faster the generations will be, and the further to the right means slower output but with greater detail.
Autonomy versus adherence to instructions: The balance between the independent creative decision-making capability of the underlying generative image model and its ability to follow or adhere to specific instructions you provide. The further the slider is to the left, the more autonomous the model is, and the further to the right means greater adherence to your prompt.
** We recommend staying in the green zone on each of the sliders **
Safe for work: Ensures that the generated images do not contain explicit, graphic, or offensive content.
Seed: Refers to the initial input or starting point used to generate an image. Unless you are getting into highly advanced prompting, you can leave this at the default value. When seed prompting, you might also notice that we inject a specific seed into the prompt form - do not change that if you want the seed prompting flow we’ve created to work as expected.
Image Sampler: Takes your prompts, samples the model, and does the work to produce the image. We’ve taken great care to curate a list of samplers that produce good results that converge.
Clip Skip: A magical key that allows you to skip layers of the model’s neural network. Clip Skip 2 is generally regarded as best practice to improve performance and convergence (see below), to help simplify the complex math running behind the scenes.
*** Convergence is important for being able to reproduce images from seed ***
Adding a new prompting form
If you accidentally delete your prompting form or want to add a new one, you can do so by going to Templates → AI Power-Ups → Image Generator. After placing the template and letting it load, you’ll get a brand new prompting form.
Seed Prompting
Seed prompting in Bluescape allows you to generate additional variations of an image that you like from previous prompt outputs. It creates an additional prompt form below the surrounding content. After generating more content, you should notice that the first image generated will be identical to the original image you Seeded from, with the exception of any changes you made to the prompt.
AI Chatbot
The AI Chatbot provides a powerful way to leverage Large Language Models to generate prompts for you based on descriptions you give it. In your AI workspaces, we have created a workflow that automates the image generation output based on the prompt recommendations of the Chatbot.
To add your Chatbot, you’ll follow the same steps as you would take to add a new prompting form: Templates → AI Power-Ups → Chat Assistant. After placing the template and letting it load, you’ll get an embedded chat interface to start prompting.
Here’s a good example prompt you can use to experience the flow - copy and paste this into the Chatbot and press send:
You are a generative image AI prompt engineer. You have a creative eye for detail and produce beautiful images and designs when asked. Please wait for instructions. You return your prompts in this format:
Prompt 1: “…”
FAQ
How long will the early access be available?
We currently don’t have a deadline for early access. If anything changes, we will email every early-access user and make an announcement here in our community.
I’m an existing Bluescape user - can I try Bluescape AI?
Absolutely. Bluescape AI is currently a separate offering from Bluescape, though we do have plans to integrate it into the core product next year. You can still sign up and try Bluescape AI with your existing account. However, if your team is using SSO, you will need to log out of Bluescape first before clicking the link to your Bluescape AI workspace. Alternatively, you can open up the link in a separate browser, in an incognito window, or in the Bluescape Desktop App (if you are using a browser) to be logged into both your main account and your AI workspace simultaneously.
If you would like to extend the trial to the rest of your organization, please contact our sales team.
Can I invite others in here with me?
Yes! You can invite as many collaborators as you want to your AI workspace. They can view your creations and generate their own if you give them the “Editor” access level. Visit our Inviting Collaborators community article for more information on inviting users.
Can I download any images I generate?
You sure can! All of the images you generate in your workspace can be easily downloaded and shared. You can download content by clicking the “Download” option when hovering over or selecting an image, or by going into the “Export” options in the “…” more menu.
What technologies are being used for Bluescape AI?
Bluescape AI is built on the Bluescape Visual Collaboration Platform and API, which has been a trusted technology used by the major studios for secure content reviews. Bluescape AI is powered by partners and technologies that you can trust, including AWS, Amazon Bedrock, Box, Adobe AEM, and Sony CI Media Cloud.
I’m a developer. How can I start building my own AI workflows in Bluescape?
Take a look at our recent Press Release to learn what we’re doing with AI. For more information on our APIs, visit our Developer Pages. If you would like to meet with our team to walk through your desired workflows and how to achieve them quickly, please reach out to bluescapeai@bluescape.com.
Some other possibilities
Model Training via Image Labeling
Metadata Handling and Integrations
Where can I get live help/support?
- Replying to this community post by using the “Reply” option below👇🏼
- Send an email to bluescapeai@bluescape.com with more information of your request.